Some observation on the data

length(kobe_data)
## [1] 29
nrow(kobe_data)
## [1] 20697
ncol(kobe_data)
## [1] 29
dim(kobe_data)
## [1] 20697    29

check if there are any null or empty values are there

##              recId        action_type combined_shot_type 
##              FALSE              FALSE              FALSE 
##      game_event_id            game_id                lat 
##              FALSE              FALSE              FALSE 
##              loc_x              loc_y                lon 
##              FALSE              FALSE              FALSE 
##  minutes_remaining             period           playoffs 
##              FALSE              FALSE              FALSE 
##             season  seconds_remaining      shot_distance 
##              FALSE              FALSE              FALSE 
##     shot_made_flag          shot_type     shot_zone_area 
##              FALSE              FALSE              FALSE 
##    shot_zone_basic    shot_zone_range            team_id 
##              FALSE              FALSE              FALSE 
##          team_name          game_date            matchup 
##              FALSE              FALSE              FALSE 
##           opponent            shot_id         attendance 
##              FALSE              FALSE              FALSE 
##         arena_temp         avgnoisedb 
##              FALSE              FALSE

Not null or missing

##              recId        action_type combined_shot_type 
##              FALSE              FALSE              FALSE 
##      game_event_id            game_id                lat 
##              FALSE              FALSE              FALSE 
##              loc_x              loc_y                lon 
##              FALSE              FALSE              FALSE 
##  minutes_remaining             period           playoffs 
##              FALSE              FALSE              FALSE 
##             season  seconds_remaining      shot_distance 
##              FALSE              FALSE              FALSE 
##     shot_made_flag          shot_type     shot_zone_area 
##              FALSE              FALSE              FALSE 
##    shot_zone_basic    shot_zone_range            team_id 
##              FALSE              FALSE              FALSE 
##          team_name          game_date            matchup 
##              FALSE              FALSE              FALSE 
##           opponent            shot_id         attendance 
##              FALSE              FALSE              FALSE 
##         arena_temp         avgnoisedb 
##              FALSE              FALSE

Transformations :

minutes remaining and second remaining can be compined to make a new variable called time remaining

plots

Factors

Graphs of all the variable with alphanumeric value.

action_type

combined_shot_type

matchup

opponent

Season

shot_type

period

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>
##     aspect: function
##     backtransform_range: function
##     clip: on
##     default: FALSE
##     distance: function
##     expand: TRUE
##     is_free: function
##     is_linear: function
##     labels: function
##     limits: list
##     modify_scales: function
##     range: function
##     render_axis_h: function
##     render_axis_v: function
##     render_bg: function
##     render_fg: function
##     setup_data: function
##     setup_layout: function
##     setup_panel_params: function
##     setup_params: function
##     transform: function
##     super:  <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>

playoffs

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>
##     aspect: function
##     backtransform_range: function
##     clip: on
##     default: FALSE
##     distance: function
##     expand: TRUE
##     is_free: function
##     is_linear: function
##     labels: function
##     limits: list
##     modify_scales: function
##     range: function
##     render_axis_h: function
##     render_axis_v: function
##     render_bg: function
##     render_fg: function
##     setup_data: function
##     setup_layout: function
##     setup_panel_params: function
##     setup_params: function
##     transform: function
##     super:  <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>

shot_zone_area

shot_zone_basic

shot_zone_range

Numerical variables that can be treated as factors

game_event_id

Continuous var

game_event_id

shot_distance

lat

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>
##     aspect: function
##     backtransform_range: function
##     clip: on
##     default: FALSE
##     distance: function
##     expand: TRUE
##     is_free: function
##     is_linear: function
##     labels: function
##     limits: list
##     modify_scales: function
##     range: function
##     render_axis_h: function
##     render_axis_v: function
##     render_bg: function
##     render_fg: function
##     setup_data: function
##     setup_layout: function
##     setup_panel_params: function
##     setup_params: function
##     transform: function
##     super:  <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>

loc_x

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>
##     aspect: function
##     backtransform_range: function
##     clip: on
##     default: FALSE
##     distance: function
##     expand: TRUE
##     is_free: function
##     is_linear: function
##     labels: function
##     limits: list
##     modify_scales: function
##     range: function
##     render_axis_h: function
##     render_axis_v: function
##     render_bg: function
##     render_fg: function
##     setup_data: function
##     setup_layout: function
##     setup_panel_params: function
##     setup_params: function
##     transform: function
##     super:  <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>

loc_y

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>
##     aspect: function
##     backtransform_range: function
##     clip: on
##     default: FALSE
##     distance: function
##     expand: TRUE
##     is_free: function
##     is_linear: function
##     labels: function
##     limits: list
##     modify_scales: function
##     range: function
##     render_axis_h: function
##     render_axis_v: function
##     render_bg: function
##     render_fg: function
##     setup_data: function
##     setup_layout: function
##     setup_panel_params: function
##     setup_params: function
##     transform: function
##     super:  <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>

lon

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>
##     aspect: function
##     backtransform_range: function
##     clip: on
##     default: FALSE
##     distance: function
##     expand: TRUE
##     is_free: function
##     is_linear: function
##     labels: function
##     limits: list
##     modify_scales: function
##     range: function
##     render_axis_h: function
##     render_axis_v: function
##     render_bg: function
##     render_fg: function
##     setup_data: function
##     setup_layout: function
##     setup_panel_params: function
##     setup_params: function
##     transform: function
##     super:  <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>

minutes_remaining

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

## <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>
##     aspect: function
##     backtransform_range: function
##     clip: on
##     default: FALSE
##     distance: function
##     expand: TRUE
##     is_free: function
##     is_linear: function
##     labels: function
##     limits: list
##     modify_scales: function
##     range: function
##     render_axis_h: function
##     render_axis_v: function
##     render_bg: function
##     render_fg: function
##     setup_data: function
##     setup_layout: function
##     setup_panel_params: function
##     setup_params: function
##     transform: function
##     super:  <ggproto object: Class CoordFlip, CoordCartesian, Coord, gg>

seconds_remaining

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

### game_date

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

attendance

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

arena_temp

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

avgnoisedb

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Multicolineariy detection using VIF

##            avgnoisedb arena_temp attendance game_date lon loc_y loc_x
## avgnoisedb          1          0          1         0   0     0     0
## arena_temp          0          1          0         0   0     0     0
## attendance          1          0          1         0   0     0     0
## game_date           0          0          0         1   0     0     0
## lon                 0          0          0         0   1     0     1
## loc_y               0          0          0         0   0     1     0
##            shot_distance lat
## avgnoisedb             0   0
## arena_temp             0   0
## attendance             0   0
## game_date              0   0
## lon                    0   0
## loc_y                  1  -1
##         Var1       Var2 value
## 1 avgnoisedb avgnoisedb     1
## 2 arena_temp avgnoisedb     0
## 3 attendance avgnoisedb     1
## 4  game_date avgnoisedb     0
## 5        lon avgnoisedb     0
## 6      loc_y avgnoisedb     0

##                      arena_temp                       game_date 
##                        81802.84                       204061.55 
##                      attendance            arena_temp:game_date 
##                       498164.70                       290160.27 
##           arena_temp:attendance            game_date:attendance 
##                       609532.82                       670177.98 
## arena_temp:game_date:attendance 
##                       788182.21
## Model :
## shot_made_flag ~ avgnoisedb + arena_temp + attendance + game_date + 
##     seconds_remaining + minutes_remaining + lon + lat + game_event_id
##        avgnoisedb        arena_temp        attendance         game_date 
##          1.350766          1.005825          1.360433          1.014508 
## seconds_remaining minutes_remaining               lon               lat 
##          1.002014          1.080305          1.001722          1.016855 
##     game_event_id 
##          1.081280

Outlier detection using cook’s D value

Relation of shot made with distance and playoffs

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.